Professional Documents
Culture Documents
Government of India
ARTIFICIAL INTELLIGENCE
USE CASE COMPENDIUM
TABLE OF CONTENTS
MOBILITY
Introduction to AI
1. Agartala: Easing its traffic management through AI and ML
2. Ahmedabad: AI helping in better Traffic Management
3. Bengaluru: Adopting AI for Automatic Number Plate Recognition
4. Jabalpur: Adopting AI to Manage Traffic and Enforcing Traffic Rules
5. Nagpur: Managing Traffic and Enforcing Traffic Rules using AI
6. Pimpri Chinchwad: Enforcing Traffic Rules by adopting AI
7. Prayagraj: Ensuring Better Traffic Management, Waste Management and Safety using AI
8. Ranchi: Adopting AI to Manage Traffic and Enforcing Traffic Rules
9. Satna: Managing Traffic and Enforcing Traffic Rules by adopting AI
10. Surat: Adopting AI for BRTS & City Bus Service.
11. Tumakuru: AI mastering the Integrated City Management Command and Control Centre
12. Varanasi: Using AI for Effective Traffic Management
13. Visakhapatnam: Effectively Managing Traffic Using AI
14. Visakhapatnam: Easing Traffic Movement Using AI
15. Kalyan Dombivali Municipal Corporation: Adopting AI to enhance Law enforcements services
WASTE MANAGEMENT
22. Varanasi: Efforts to make the city cleaner using AI
23. Visakhapatnam: Creating a Healthy Metropolis using AI
EDUCATION
24. Belagavi: Adopting AI to Enable the Digital Education
25. Pimpri Chinchwad: Adopting AI to Measure the Education Effectiveness in Municipal E-Classrooms
TABLE OF CONTENTS
E-GOVERNANCE
26. Bengaluru: Adopting AI to enhance Law enforcements services
27. Bengaluru: Adopting AI for Sentiment Analysis of the citizens
28. Jaipur: AI plays a major role in the effective governance system
29. Pune: AI boosting efficiency in Property Tax Assessment
ENERGY
30. Bhopal: Installs AI-powered Smart Poles Across the City
GLOSSARY
AI Artificial Intelligence
AR Augmented Reality
ML Machine Learning
VR Virtual Reality
INTRODUCTION TO AI
The ever-expanding use of Artificial Intelligence (AI) and machine learning (ML) in various
sectors has triggered interest and numerous opportunities for governments across the
world. By deploying AI in the public sector, traditional ways of working can be changed
extensively to bring in more efficiency and effectiveness in service delivery. AI has the
potential to make sense of the vast data being collected and generate actionable insights.
This can help improve the performance of cities, optimize operational costs and resources
and enable proactive citizen engagement, thus enhancing quality of life of the citizens.
The Ministry of Housing and Urban Affairs has spearheaded the establishment of
significant building blocks for data, along with robust governance structures and toolkits
to enable their success. Thus, it is best placed to leverage the work of these interventions
through the thriving ecosystem partnerships to dovetail into exploring the use of
emerging tech solutions to address the requirements of the ever-expanding needs of
urban India.
Recognizing the value of AI, the Smart Cities Mission at the Ministry of Housing and Urban
Affairs is developing an AI Playbook for Cities to enable cities to utilize the power of AI for
urban governance, thus making them more efficient, liveable and sustainable. It
emphasizes on how cities may operationalize this technology to improve the overall city
functioning and service delivery to citizens.
It is heartening to note that some of the smart cities have already begun deploying AI
solutions in their cities to good effect. Thus, in addition to the AI Playbook, this AI Use Case
Compendium will help cities learn from each other and follow the best practices
implemented so far. This document highlights the role of AI as a solution to address the
pervasive challenges in Indian cities. The use cases are spread across various urban
domains like Mobility, Safety and Security, Waste Management, Education, E-Governance
and Energy.
MOBILITY
1 AGARTALA
Easing its traffic management through
AI and ML
02
Key Highlights of the implementation
The traffic density AI algorithm was finalized and arrived at through behavioral study of traffic vehicle density
across all the 22 junctions at different times of the day.
The AI algorithm was optimized through predictive ML (Machine Learning process) using historical traffic
density data resulting in inefficient traffic movement across these junctions.
Support Ecosystem: MSI - Honeywell Automation India Ltd, Consultant –Price water house Coopers Pvt
Ltd, Traffic Department, Govt. of Tripura
Scale of Deployment
This ITMS intervention has been done for 22 junctions across the city of Agartala.
Traffic department has plans to further scale up the scope of deployment including – number of junctions,
area, units etc.
03
1.4 Challenges faced during Implementation
• One of the challenges faced during the implementation was narrow roads with minimum carriage way.
• Secondly decision on positioning of the virtual rectangular loops which was monitored by ‘Flir” camera.
• Human factor was the major challenge during the implementation of ITMS adaptive and predictive
solution since vehicle owners were not accustomed to stop at designated stop lines at traffic junctions of
Agartala city. This human action resulted in incorrect vehicle density being captured by “Flir” camera with
a corresponding wrong input pulse being fed to the “Onnyx” Controller which initially resulted in long
wait time for vehicles at these junctions. This was mitigated through stringent action taken by Traffic
Department; Govt. of Tripura wherein stop-line outliers were penalized, and fines were imposed.
72.0
Post Go-Live of this solution at the junction in October 2020 it was observed that there was a significant
improvement in average wait time with a minimum of 8 seconds and a maximum of 15.3 seconds of wait time
as seen in Graph-II.
15.3 15.3
14.8
04
1.6 Key Learnings from Deployment
• One of the challenges faced during the implementation was narrow roads with minimum carriage way.
• Secondly decision on positioning of the virtual rectangular loops which was monitored by ‘Flir” camera.
• Human factor was the major challenge during the implementation of ITMS adaptive and predictive
solution since vehicle owners were not accustomed to stop at designated stop lines at traffic junctions of
Agartala city. This human action resulted in incorrect vehicle density being captured by “Flir” camera with
a corresponding wrong input pulse being fed to the “Onnyx” Controller which initially resulted in long
wait time for vehicles at these junctions. This was mitigated through stringent action taken by Traffic
Department; Govt. of Tripura wherein stop-line outliers were penalized, and fines were imposed.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
05
2 AHMEDABAD
AI helping in better Traffic
Management
• 92 junctions are being made zero tolerance for red light violation.
• Red Light Violation Detection and Automatic Number Plate Recognition.
• Integration with VAHAN portal of RTO.
• Data collected from 6000 surveillance cameras including ANPR and RLVD.
Support Ecosystem: City Traffic Police Department as a key user department, SCADL, PMC
Tech Providers: M/s. Trimax Smart Infraprojects Pvt. Ltd.
Scale of Deployment
Currently the system is installed at 92 junctions with data from 6000 surveillance cameras.
06
2.4 Challenges faced during Implementation
• Making Citizens to be aware about the system and their acceptance for the challan generated through
the system.
JS
</>
{ .. .}
P
PH
;
D E}
O
L {.C
XM
+
C+
07
3 BENGALURU
Adopting AI for
Automatic Number Plate Recognition
• The ICCC system should be integrated to receive alarms/alerts from Vendor for the flagged vehicle
(with number tagged as “Suspicious”, “Stolen”) across the city on map with relevant alert.
• The dashboard has been planned to be created.
Support Ecosystem: Pentaho Enterprise Edition, Bengaluru smart city Ltd. (BenSCL).
Consultant: iDeck, BTP department.
Scale of Deployment
08
3.4 Challenges faced during Implementation
• Fetching data from other stakeholder departments
• Managing the server space to store the captured events by the system
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
09
4 JABALPUR
Adopting AI to Manage Traffic and
Enforcing Traffic Rules
a. Traffic signal time is automatically adjusted through the Intelligent traffic Control system at the
junction (a set of hardware and software).
b. Improved traffic/congestion management in the city during peak hours and reduction in the travel
time of the commuters.
c. City traffic police to identify and track the violations for the necessary issuance of challans through an
Automatic number plate detector Camera, Speed Violation & Red-light violation detection cameras.
d. Enforcing the traffic rules remotely from the command centre while getting real-time camera feed of
the junction.
10
4.3 Implementation Process
The project was implemented through RFP process and it was for full deployment without any POC.
Support Ecosystem: JP Softek, Traffic Police, M/s PwC (PMC), Jabalpur Smart City Ltd, Nagar Nigam, and
Postal Department
Scale of Deployment
11
4.6 Key Learnings from Deployment
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
12
5 NAGPUR
Managing Traffic and
Enforcing Traffic Rules using AI
• Improved traffic/congestion management in the city during peak hours and reduction in the travel
time of the commuters.
• City traffic police to identify and track the violations for the necessary issuance of challans through an
Automatic number plate detector Camera, Speed Violation and Red-light violation detection cameras.
• Enforcing the traffic rules remotely from the command centre while getting real-time camera feed of
the junction.
13
5.3 Implementation Process
RLVD systems functioning is based on four cameras working as a single unit. Out of the four cameras, one
camera is used for capturing the status of the signal (i.e., Red/Green/Orange) and captures the event as proof
of violation. The other three cameras are used for capturing the Number plate of the vehicles jumping the
Red signal.
The project was implemented through RFP process and it was for full deployment without any POC.
Support Ecosystem:Traffic Police Department Nagpur, Nagpur Smart and Sustainable City Development
Corporation Limited (NSSCDCL), Nagpur Municipal Corporation (NMC), SI M/s Larsen and Toubro Limited.
Tech Providers: M/s L&T is Master System Integrator & Service Provider
Scale of Deployment
14
5.6 Key Learnings from Deployment
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
15
6 PIMPRI CHINCHWAD
Enforcing Traffic Rules by
adopting AI
The movement of vehicles is monitored with respect to the signalling at junctions. Status of signals is
captured as a proof of violations, along with the vehicle number captured.
16
6.3 Implementation Process
A Proof of Concept (POC) is being conducted at selected 10-12 locations. Based on real-time alarms
generated & analysis of the alarms, results are being improved at selected POC sites. The rolloutAI-
basedased analytics at remaining camera locations shall be done subsequently.
Scale of Deployment
City Surveillance and Traffic Cameras across identified areas of the city:
• No of Locations: 606
• No of Fixed Box Cameras: 1520
• No of ANPR Cameras: 850
• No of AI Licenses: 1000 (Dynamic)
17
6.6 Key Learnings from Deployment
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
18
7 PRAYAGRAJ
Ensuring Better Traffic Management,
Waste Management and Safety using AI
SWM/Vandalism -
• Identifying the places/ areas with graffiti for further cleaning and preventive action plan to minimise
such activity and help to keep the area clean.
• Real-time vandalism detection by monitoring object activities and take the prompt action against such
activity.
Debris and Garbage detection -
• Detection of debris and garbage cleaning activity initiated by the municipal team as per define SOP.
Traffic Management -
• Detection of Parking violations, speeding vehicles, Wrong way, or illegal turns
• Identify the areas where such violation is trending.
• Based on the report traffic police plan the enforcement for action.
Citizens Safety -
• Detecting the suspicious person/ activity at sensitive locations (Loitering detection).
• Controlling the inflow/outflow of the public in case of large gatherings.
19
7.3 Implementation Process
AI Based Intelligent Video Analytics solution is designed in a way to create alerts on decisive events as per the
scope of the analytics requirement. Once the area is identified, post that CCTV cameras are placed at the right
positions to capture the events.
Support Ecosystem: Prayagraj smart city, Municipal corporation, Department of police Project
management team.
Tech Providers: Larsen & Toubro Ltd, Construction - Smart World & Communication
Scale of Deployment
20
7.5 Impact created on City
• Better monitoring of the Traffic for smooth flow and safe commute.
• Improved AI Based Intelligent Video Analytics service through efficient staff deployment.
• Reduction in Accidents and Traffic Jams by continuously monitoring the flow.
• Improved safety for the citizens by reduction in the crime rate.
• Efficient monitoring of the cleanliness in the city by deploying waste management analytics.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
21
8 RANCHI
Adopting AI to Manage Traffic and
Enforcing Traffic Rules
• Al-based Automatic Number Plate Recognition (ANPR) cameras and software used to provide vehicle
identification or detect crimes and violations in real time.
• Al based Red Light Violation Detection (RLVD) system is fully automated mass surveillance system
identifying vehicles jumping the red light, stopping after the red-light mark, and over the zebra-
crossing.
• Implementation of Speed Violation Detection (SVD) system and integrated e-Challan to identify
vehicles not adhering to the permissible speed limits on the roads, is very robust.
• AI - Adaptive Traffic Control System (ATCS) automatically adapts the timings of traffic lights based on
real-time traffic conditions to optimize the flow of traffic.
22
8.3 Implementation Process
The project was implemented through RFP process and it was for full deployment without any POC.
Support Ecosystem: State Government, Transportation Departments, RMC and Ranchi traffic police
Tech Providers: Honeywell Automation India Limited
Scale of Deployment
The city plans to develop more junctions by installing the system equipment. Upgradation of six locations
out of 81, development of smart lighting, solid waste management processes, and a business model for C4
that uses VMSB and PA are envisioned.
+
C+
23
9 SATNA
Managing Traffic and Enforcing Traffic Rules
by adopting AI
• Cameras installed were being used for surveillance and monitoring traffic behaviour across various
locations and E-challan is generated for validated violations.
• ITMS cameras come in handy for police to check crimes. The camera captures video of high resolution
which makes it easy to identify people and do live view surveillance.
24
9.3 Implementation Process
ITMS Project was done by Satna Smart City through an RFP process. After the finalization of the bidder, the
project is implemented in Satna city.
Execution process is described as below:
Stage 1- Completion of Site Survey.
Stage 2- Project Inception.
Stage 3- Delivery and Receipt of Hardware and Software at site (Warehouse) and after verification of such
items by SSCDL.
Stage 4- Power-up (for hardware), Installation, Configuration and Application deployment, Integration and
Go-Live.
Stage 5- Completion of Integration requirements.
Stage 6- Dry Run and Go-Live.
Scale of Deployment
25
9.5 Impact created on City
• Ranchi Traffic Management Systems are solutions that municipalities integrate into traffic cabinets and
intersections, to make fast, cost-effective improvements in traffic.
• NPR, ATCS, PA, ECB, and VMSB are all used at Ranchi deployment Smart Traffic Engineering to ensure
fast, free and efficient traffic with fewer accidents.
• Mass surveillance system is identifying vehicles jumping the red light, stopping after the red-light mark,
and over the zebra-crossing.
• AI - Adaptive Traffic Control System (ATCS) is a resolve that automatically adapts the timings of traffic
lights based on real-time traffic conditions to optimize the flow of traffic.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
26
10 SURAT
Adopting AI for BRTS &
City Bus Service
27
10.3 Implementation Process
The project was implemented through RFP process and it was for full deployment without any POC.
Support Ecosystem: NEC India Pvt. Ltd and ARS T&TT Pvt. Ltd
Tech Providers: Lumiplan India ITS Pvt. Ltd
Scale of Deployment
The software was used for a total of 779 fleet of buses for the entire city.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
28
11 TUMAKURU
AI mastering the Integrated City Management
Command and Control Centre
29
Video Analytics (VA) strengthens the traffic management systems for improved security by using
metadata. The purpose of a surveillance video-based system designed for traffic analytics is to detect,
track, and analyse vehicle patterns and behaviour concerning abnormal event predication and detection
on the roads as well as the crimes. The rapid growth in the area of Internet of Things (IoT), big data, and
analytics has drastically changed the Intelligent Transportation System (ITS) over the years. There has
been an increase in the rising concerns for proper traffic management due to traffic congestion, road
accidents, potential threats, excessive pollution levels, etc. Smart solutions like VA help cater to such
problems and assist cities, pedestrians, and officials in addressing these growing concerns. Through
surveillance cameras, the police can both prevent crimes from happening and can quickly solve criminal
cases with material evidence. In addition, surveillance cameras protect against property theft, and
vandalism. It is very difficult to get away with stealing something if there are cameras filming. Through
this Surveillance we can monitor the city Activity in real time in a Remote place. Through Video Analytics
we can easily detect, track, and analyze moving or stationary objects.
Support Ecosystem: Police Department and Tumakuru Smart City Corporation Ltd
Scale of Deployment
30
11.5 Impact created on City
• Total Challans worth Rs 15.00 lakhs has been issued to the violators using automated system.
• 20 Incidents of garbage thrown on the road and important places are anatomically identified.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
31
12 VARANASI
Using AI for
Effective Traffic Management
• Detecting congestion on road and generating real time alerts, so that the concerned authorities can act
upon it.
• Monitoring of congestion is done based on pre-historic data based on the traffic volume in different
time periods of the day.
• Real-time announcements through PA System.
32
12.3 Implementation Process
An RFP was floated, and full deployment is done for the implementation of the solution.
Scale of Deployment
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
33
13 VISAKHAPATNAM
Effectively Managing Traffic Using AI
34
13.3 Implementation Process
Greater Visakhapatnam Smart City Project envisages deployment of following components to achieve the
objectives:
• Deployment of various sensors (cameras, traffic violation, environment and weather sensors) throughout
the city to improve situational awareness.
• Development of command and communication centre for improved visualization of the ambient
situation in the city and facilitation of data driven decision making. The intent of this RFP is to invite bids
from the Bidders for the implementation of an integrated solution.
The Request for Proposal (RFP) consists of three volumes as detailed:
RFP Volume 1: Instruction to Bidders.
RFP Volume 2: Scope of work including Functional & Technical Specifications.
RFP Volume 3: Master Service Agreement.
The contract was awarded to L&T in consortium with Fluentgrid Limited after the successful bidding.
Scale of Deployment
• RLVD system was implemented at 10 major junctions across the city and 65 lanes were under continuous
surveillance.
• More than 100 RLVD and ANPR cameras are installed and monitored on daily basis for any violations.
Automatic challan generation option was implemented which enables the system to send a notification
to the owner of the vehicle along with the type of violation and the details captured while violating the
signal.
35
13.6 Key Learnings from Deployment
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
36
14 VISAKHAPATNAM
Easing Traffic Movement Using AI
Apart from the signalling based on vehicle density, ATCS can also be operated in the following mode:
• Manual- The software was integrated to the COC application and the traffic signals can be controlled
manually.
• VIP mode- This ensures the smooth flow of the emergency vehicles like ambulances to skip the signals
to reach the destination faster which in turn improves the response time for the patient.
37
• Development of command and communication centre for improved visualization of the ambient
situation in the city and facilitation of data driven decision making. The intent of this RFP is to invite bids
from the Bidders for implementation of an integrated solution for implementation of an integrated
solution
The Request for Proposal (RFP) consists of three volumes as detailed:
RFP Volume 1: Instruction to Bidders.
RFP Volume 2: Scope of work including Functional & Technical Specifications.
RFP Volume 3: Master Service Agreement.
The contract was awarded to L&T in consortium with Fluentgrid Limited after the successful bidding.
Scale of Deployment
50 major junctions were identified in the city and ATCS was deployed at these junctions.
JS
</>
{ .. .}
P
PH
;
D E}
O
L {.C
XM
+
C+
38
SAFETY
AND
SECURITY
15
KALYAN DOMBIVALI MUNICIPAL
CORPORATION
Adopting AI to enhance Law enforcements
services
Key highlights of the AI-enabled crime prediction & police presence optimization as-a-Service are
as follows:
• The AI solution utilizes advanced hot spot identification models, risk terrain analysis, regression,
classification, and clustering models for identifying and analysing patterns and trends in crime,
enabling police presence optimization.
40
• Advanced crime-mapping tools that integrate with public safety agency databases, IoT sensors and
surveillance systems to generate crime locations and recommend the best strategy to provide
effective safety services and optimize savings on resources and time.
Support Ecosystem
Master System Integrator (MSI): Consortium of NEC Corporation India (Lead Bidder), Allied Digital Services
and CMS Computers
Consultant: PricewaterhouseCoopers Pvt. Ltd. (PwC India)
Key Users/Stakeholders: Kalyan Dombivli Municipal Corporation (KDMC) and its departments, Maharashtra
Police & Traffic Department
Nodal Organization for Implementation: Smart Kalyan Dombivali Development Corporation (SKDCL)
Tech Providers:
OEM: Trinity
Scale of Deployment
The total scope is as follows:
• CCTV Locations – 301 Locations
• CCTV Cameras (including Fixed Box Cameras, PTZ
• Cameras and Exit ANPR Cameras) – 877 Nos.
• Dashboard Cameras: 36 Nos.
• ITMS Junctions (along with RLVD and ANPR): 13 Nos.
• Environment Sensors: 13 Nos
41
15.4 Challenges faced during implementation
• As this is still under implementation, the challenges are in the process of being consolidated.
• Challenges faced during implementation will be shared once the implementation is completed
JS
</>
{ .. .}
P
PH
;
D E}
O
L {.C
XM
+
C+
42
16 NEW TOWN KOLKATA
Mitigating COVID-19 impact
by adopting AI
43
Key Highlights of the implementation
• AI technology with ML, Deep learning and Computer vision techniques were adopted.
• The audio alert enables enforcement of covid norms related to the wearing of masks.
Support Ecosystem
MSI- NIS Facility Management Services Pvt Ltd; Urban Local Body- New Town Kolkata Development
Authority; Smart City- New Town Kolkata Green Smart City Corporation Limited
Tech Providers:
Videonetics Pvt Ltd
Scale of Deployment
The solution has only been deployed at one community market i.e., CB Market.
44
16.4 Challenges faced during implementation
• Due to the 2nd wave of the Covid-19 Pandemic, mask wearing was made mandatory at NKDA area. The
challenge was to detect specific cases accurately in a high footfall environment and finding the exact spot
where the cameras can be installed.
• Creating a sense among the people that they are being monitored and help arrest the daily infection level
is quite a challenge. The audio message had to be clear, short, polite but persuasive and capable of
creating a sense among the people that they are being monitored and help arrest the daily infection
level.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
45
17 PIMPRI CHINCHWAD
Ensuring Traffic Safety using AI
• Every Vehicle Number Plate will be captured under Automated Number Plate Recognition, AI feature.
46
17.3 Implementation Process
A Proof of Concept (POC) is being conducted at selected 10-12 locations. Based on real time alarms
generated & analysis of the alarms, results are being improved at selected POC sites. The rollout of AI based
analytics at remaining camera locations shall be done subsequently.
Support Ecosystem
MSI: Tech Mahindra led Consortium
Consultant: Ernst & Young LLP
Client: PCSCL & PCMC IT & Telecom
Tech Providers:
Videonetics Pvt Ltd
Scale of Deployment
City Surveillance and Traffic Cameras across identified areas of the city:
• No of Locations: 606
• No of Fixed Box Cameras: 1520
• No of ANPR Cameras: 850
• No of AI Licenses: 1000 (Dynamic)
47
17.5 Impact created on City
Improvement in correct alarm generations and reduction in false alarms. Addressal of correct alarms
effectively in short duration.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
48
18 PIMPRI CHINCHWAD
Monitoring City’s Security using AI
• Alert is raised upon vehicle intrusion and identification of suspicious objects within the surveillance
area.
• Controlled public premises ensuring better safety and security.
Support Ecosystem
MSI: Tech Mahindra led Consortium
Consultant: Ernst & Young LLP
Client: PCSCL & PCMC IT & Telecom
Tech Providers:
Videonetics Pvt Ltd
49
Scale of Deployment
City Surveillance across identified areas of the city:
• No of Locations: 606
• No of Fixed Box Cameras: 1520
• No of Fixed Box Camera along with FRS: 30
• No of Dome Cameras: 383
• No of PTZ Cameras: 160
• No of AI Licenses: 4816 (Dynamic)
50
18.5 Impact created on City
Improvement in correct alarm generations and reduction in false alarms. Addressal of correct alarms
effectively in short duration.
JS
</>
{ .. .}
P
PH
;
D E}
O
L {.C
XM
+
C+
51
19 VARANASI
Ensuring the city’s security by adopting AI
The framework comprises stream and batch processing capabilities. Every component of the Analytics
layer, Message Broker, Streaming, NoSQL, and Search Indexer can be horizontally scaled. The streaming
analytics pipeline can be used for processes like anomaly detection, alerting, and computation of statistics
like traffic flow rate. Batch processing can be used to extract patterns in the data, look for anomalies over
a period of time, and build machine learning models. The data is kept in a NoSQL database for state
management, e.g., the occupancy of a building, activity in a store, or public movement in a train station.
This also provides the capability for forensic analytics, if needed. The data can be indexed for search and
time-series analytics. The information generated by streaming and batch processing is exposed through a
standard API for visualization. The API can be accessed through REST, WebSocket, or messaging, based
on the use case. The user interface allows the user to consume all the relevant information.
• Alert is raised to the concerned authorities if any person/vehicle/animal intrudes the area under
supervision.
52
19.3 Implementation Process
An RFP was floated, and full deployment is done for the implementation of solution.
Support Ecosystem
Varanasi Police Team
Tech Providers:
Efkon India Pvt Ltd
Scale of Deployment
It is currently deployed in the defence land area, cantonment sector, police headquarters entry/exit
locations.
53
19.6 Key learnings from Deployment
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
54
20 VARANASI
AI in Monitoring of Public Safety
during Pandemic
The framework comprises stream and batch processing capabilities. Every component of the Analytics
layer, Message Broker, Streaming, NoSQL, and Search Indexer can be horizontally scaled. The streaming
analytics pipeline can be used for processes like anomaly detection, alerting, and computation of statistics
like traffic flow rate. Batch processing can be used to extract patterns in the data, look for anomalies over
a period of time, and build machine learning models. The data is kept in a NoSQL database for state
management, e.g., the occupancy of a building, activity in a store, or public movement in a train station.
This also provides the capability for forensic analytics, if needed. The data can be indexed for search and
time-series analytics. The information generated by streaming and batch processing is exposed through a
standard API for visualization. The API can be accessed through REST, WebSocket, or messaging, based
on the use case. The user interface allows the user to consume all the relevant information.
• Alert is raised to the concerned authorities if the number of people exceed the normal count and an
announcement using PA/VaMS is made for crowd disbursement in such areas.
55
20.3 Implementation Process
An RFP was floated, and full deployment is done for the implementation of solution.
Support Ecosystem
Varanasi Police Team
Tech Providers:
Efkon India Pvt Ltd
Scale of Deployment
This is deployed during sensitive times when government imposes restrictions on public gatherings during
Covid lockdown, riot times etc.
56
20.6 Key learnings from Deployment
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
57
21 VISAKHAPATNAM
Making the city a safer place to
live using AI
AI was used to monitor any untoward incidents such as overcrowding or unattended baggage. The live
video feed of the cameras is processed through Video Management System (VMS) which detects the
overcrowding or the unattended baggage. After detection, an alert will be triggered in the COC
application.
An alert mechanism along with a Standard Operating Procedure is configured in the COC application. All
the concerned officials will be notified through SMS and email from the COC application.
58
Key Highlights of the implementation
• Overcrowding Alert: The police department operator will check the reason for the gathering and
ensure that the place was evacuated. The COC operator will continuously monitor the surveillance till
the overcrowding was cleared. The Standard Operating Procedure (SOP) followed by ICCC for an
Overcrowding Alert is shown in the figure below.
• Unattended Baggage Alert: The police department operator will check the reason for the gathering
and ensure that the place was evacuated. The COC operator will continuously monitor the
surveillance till the overcrowding was cleared. The Standard Operating Procedure (SOP) followed by
ICCC for an Overcrowding Alert is shown in the figure below.
• Bomb Identification Alert: In case of the bomb identified alert, the COC operator alerts the citizens
through Public Addressing System (PAS) and Variable Messaging Displays (VMD’s). The operator will
notify the fire department and ambulance service along with monitoring the situation through
surveillance. Once the situation is taken under control, the alert will be closed. The Standard
Operating Procedure (SOP) followed by ICCC for an Overcrowding Alert is shown in the figure below.
Support Ecosystem
IMT Operator
Police department and
Traffic division, Fire Department, EMRI Service
Tech Providers:
L & T pvt. ltd. and Fluentgrid Consortium
59
Scale of Deployment
525 cameras are installed across 83 major junctions the city. Out of which 100 are PTZ (Pan Tilt Zoom
Cameras). 70 extra cameras were deployed after the implementation of the project. The scalability is planned
to be increased based on the future requirements.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
60
WASTE
MANAGEMENT
22 VARANASI
Efforts to make the city cleaner
using AI
The framework comprises stream and batch processing capabilities. Every component of the Analytics
layer, Message Broker, Streaming, NoSQL, and Search Indexer can be horizontally scaled. The streaming
analytics pipeline can be used for processes like anomaly detection, alerting, and computation of statistics
like traffic flow rate. Batch processing can be used to extract patterns in the data, look for anomalies over
a period of time, and build machine learning models. The data is kept in a NoSQL database for state
management, e.g., the occupancy of a building, activity in a store, or public movement in a train station.
This also provides the capability for forensic analytics, if needed. The data can be indexed for search and
time-series analytics. The information generated by streaming and batch processing is exposed through a
standard API for visualization. The API can be accessed through REST, WebSocket, or messaging, based
on the use case. The user interface allows the user to consume all the relevant information.
• Generates a real-time alert to the concerned authorities like Municipal Corporation, SWM team,
Nagar Nigam through E-Mail, SMS etc.
62
22.3 Implementation Process
• An RFP was floated, and full deployment is done for the implementation of solution.
Support Ecosystem
Varanasi Smart City
Tech Providers:
Efkon India Pvt Ltd
Scale of Deployment
525 cameras are installed across 83 major junctions the city. Out of which 100 are PTZ (Pan Tilt Zoom
Cameras). 70 extra cameras were deployed after the implementation of the project. The scalability is planned
to be increased based on the future requirements.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
63
23 VISAKHAPATNAM
Creating a Healthy Metropolis
using AI
They are designed to ensure that there will be no spillage while throwing the garbage or while lifting the
garbage. To adhere to the Swachh Bharat principles of source segregation, two separate bins for wet and
dry waste were deployed.
64
Key Highlights of the implementation
• 50 such bins were deployed across the city with a 1-ton capacity volume.
• Automated and on-demand servicing using volumetric sensors.
• Rule based alerts.
• Proactive tracking of 525 vehicles and dynamic vehicle route allocation.
• Missed bin alert.
• Real time weight recording at the transfer station
Support Ecosystem
SWM Operators, PHD
Garbage Collection teams
Tech Providers:
L & T pvt. ltd. and Fluentgrid Consortium
Scale of Deployment
65
23.4 Challenges faced during implementation
• Behavioural change in citizens for source segregation is the major challenge faced during
implementation.
• Dynamic route allocation of vehicle to optimize the garbage collection efficiency was a hurdle while
implementing the process.
JS
</>
{ .. .}
P
PH
;
D E}
O
L {.C
XM
+
C+
66
EDUCATION
24 BELAGAVI
Adopting AI to Enable
the Digital Education
Using Deep Learning and Analytics, the mood, the reading enthusiasm, and the interest areas were
captured. Thousands of lines of codes sit inside the eBooks (EPUB Standard supports active code inside
the eBooks and is supported on Windows, Linux, Mac, Android, and iOS), which evaluates the reader's
behaviour and feeds it to the AI algorithm. Based on the suggestion from the AI algorithm, the story
paragraphs dynamically change to the reader's interest.
It also changes the vocabulary of the future pages, basis the student's lookup for dictionary words to
make it easily readable. The same is implemented using Active eBooks for competitive examinations in a
way that the eBook automatically gives more subtle questions using different cognitive techniques based
on proficiency analytics. The cognitive deep learning algorithm will send the data to an AI learning
algorithm seated inside the eBooks.
68
Key Highlights of the implementation
Support Ecosystem
PMC-MSI-Ostillio Systems Pvt. Ltd, Belagavi Smart City.
Tech Providers:
Johan Brown Pvt Ltd and Nestor Analytics
Scale of Deployment
• At present, it is one e-Library with 3 floors. Now the deployment is being scaled for Pre-Primary and
primary students in their cognitive learning.
69
24.4 Challenges faced during implementation
• Learning data is individual driven in most cases. 90% of the users visit the library for exam preparation.
• Data gathering for competitive exams is a time-consuming process and is being done rapidly.
• Initially, public and avid readers were sceptical about the value addition that AI brings in until they
evidenced it at the e-library
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
70
25 PIMPRI CHINCHWAD
Adopting AI to Measure the Education
Effectiveness in Municipal E-Classrooms
Students Behaviour:
• Estimation of students perceived attention in the classroom.
• Classroom behaviour analysis: run, danger, fight, fall.
• Student mood and emotion analysis.
• Attendance with a timestamp of entry, exit, and duration of presence inside the class for students.
Teacher Behaviour:
• Effective utilization of smart board and black/white board.
• Time spent engaging with the students.
• Time spent idle, on phone, sitting, chatting with other staff etc.
• Attendance for teachers and regularity to the classes.
• Punishment and harassment analytics.
• Two cameras were installed in every classroom. One among the two cameras was installed above the
blackboard/ smartboard whereas the other one was placed on the opposite wall of the first camera.
• The camera installed above the blackboard/ smartboard captures students’ and another one captures
student and teachers’ data respectively. The captured data was transmitted to a server where AI
analytics software was installed. Subsequently, analytics were performed, and reports were
generated.
• Currently, due to the covid guidelines, teachers and students cover their faces by masks. Hence, the
face recognition feature is not enabled. This will be implemented in all the schools once the
restrictions are lifted.
71
25.3 Implementation Process
AI based education analytics was a project component of municipal e classroom. Currently, this has been
implemented on pilot basis. The bidder was selected through an open competitive bidding process to
implement the project.
Support Ecosystem
Education Department & IT department PCMC
Consultant: KPMG Advisory Services Pvt Ltd.
Tech Providers:
Bennett Coleman & Co. Ltd. in consortium with Edique Solutions Pvt Ltd
Scale of Deployment
• At present, the education analytics were carried out in 3 schools and 18 Classrooms.
72
25.6 Key learnings from Deployment
JS
</>
{ .. .}
P
PH
;
D E}
O
L {.C
XM
+
C+
73
E-GOVERNANCE
26 BENGALURU
Adopting AI to enhance Law
enforcements services
• Accurate Chatbot functionality has been integrated into Bengaluru Smart City Web Portal.
Deployment into Bengaluru Smart City mobile app and Open Data Portal is underway.
• A citizen can either get information or register a grievance and get it redressed without human
intervention.
Support Ecosystem
Azure, Bengaluru smart city Ltd. (BenSCL).
Infrastructure Development Corporation (Karnataka). iDeck
Tech Providers:
Fluent Grid
Scale of Deployment
75
26.4 Challenges faced during implementation
• Gathering multi department data to get it disseminated.
JS
</>
{ .. .}
P
PH
;
D E}
O
L {.C
XM
+
C+
76
27 BENGALURU
Adopting AI for Sentiment Analysis
of the citizens
• The ICCC system will be integrated with the Solution and provides analytics based on the social
media feed collected from the open social data from Facebook, Twitter etc.
• Gathering and generating actionable items by raising the incidents in real time.
Support Ecosystem
Bengaluru smart city Ltd. (BenSCL).
Consultant: iDeck,
Social Media handles of line departments integrated with ICCC.
Tech Providers:
Fluent Grid
77
Scale of Deployment
JS
</>
{ .. .}
P
PH
;
D E}
O
L {.C
XM
+
C+
78
28 JAIPUR
AI plays a major role in the effective
governance system
79
Support Ecosystem
Jaipur Nagar Nigam (Heritage)/Jaipur Smart City Limited
Consultant: - Egis India Consulting & Engineers Ltd
Tech Providers:
Fluent Grid
Scale of Deployment
• At present, the solution is deployed in all the ward offices of Jaipur Nagar Nigam Heritage.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
80
29 PUNE
AI boosting efficiency in Property
Tax Assessment
Support Ecosystem
SI - M/s. La Mere Business Private Ltd,
Consultant- Ernst and Young LLP,
Municipal teams – IT Department PMC, Property Tax Department, PMC.
Tech Providers:
Fluent Grid
Scale of Deployment
• At present, the solution is deployed at PMC Jurisdiction, initially started with 2 wards of PMC.
81
29.4 Challenges faced during implementation
• One of the major challenges was gathering of requirements from the department, data availability to
apply the AI, domain knowledge gaps at the time of RFP preparation, rigid policy of various departments
to share the data.
• Integration of property tax survey app data and
• citizen data is challenging.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
82
ENERGY
30 BHOPAL
Installs AI-powered Smart Poles
Across the City
• Using the New age technologies such as AI/ML, IoT these poles have proven to be beneficial in
helping the city administration effectively provide situation awareness across domains like Public
Transport, Environment, Solid Waste Collection, etc.
• The smart pole solution comprises a network of key infrastructure elements with a combination of
sensors and peripheral devices that can be managed through the internet to monitor and control
streetlights.
• Object Detection - Animals, Humans, Cars, Trucks, Buses
• Intuitive backend dashboard to manage content and perform trigger analytics for ROI optimization.
84
30.3 Implementation Process
As per the terms of RFP, Smart Pole consists of Environmental Sensors, Wi-Fi, Cameras, Smart LED Street
Lights, EV Charging Point. All these components are further connected and integrated with the ICCC where
operators regularly monitor Traffic situations, Environmental conditions, Conditions of the Streets lights, and
using further technological interventions the data collected from there is processed and analysed to provide
city administration with actionable inputs to act upon.
Support Ecosystem
Master System Integrator (MSI):
Implementation Bhopal Smart City Development Corporation Limited, Ericsson, BMC
Tech Providers:
OEM: Bharti Infratel
Scale of Deployment
• Smart pole installation across the city: 150Smart Poles installed and 400 Smart poles to be installed.
• LED street light installation: 20,000 LED streetlights to be installed by replacing the conventional sodium
lamps and mercury lamps.
• 87 Smart Poles are deployed with Environment sensors.
85
30.5 Impact created on City
• The Implementation of LED Street lights helped improve the quality of life of city residents by improving
the city lighting.
• Smart pole has Li- ion batteries to eliminate diesel generator as a secondary power source. Li-ion battery
provides the back up during electricity outage.
• Optical fibre networks across the city ensure robust connectivity and enable the city to accommodate
future technologies. It will also help to establish connectivity between Government departments, City
infrastructure, and Command & Control Centre.
• 4- 46% per year energy saving.
• 98% uptime will help to reduce carbon footprint.
• Wi-Fi connectivity at 100 Hot Spots locations where city resident footfall is increasing. It provides free Wi-
Fi for defined time periods to the registered users
• Camera surveillance ensures safety, security, and parking management in the city.
JS
</>
{ .. .}
P
PH
};
DE
O
L {.C
XM
+
C+
86
Ministry of Housing and Urban Affairs
Government of India